Using Object Relational Mapping with ActiveRecord


You want to store data in a database without having to use SQL to access it.


Use the ActiveRecord library, available as the activerecord gem. It automatically defines Ruby classes that access the contents of database tables.

As an example, lets create two tables in the MySQL database cookbook (see the chapter introduction for more on creating the database itself). The blog_posts table, defined below in SQL, models a simple weblog containing a number of posts. Each blog post can have a number of comments, so we also define a comments table.

	use cookbook;

	CREATE TABLE blog_posts (
	 title VARCHAR(200),
	 content TEXT,

	CREATE TABLE comments (
	 blog_post_id INT(11),
	 author VARCHAR(200),
	 content TEXT,

Here are two Ruby classes to represent those tables, and the relationship between them:

	require cookbook_dbconnect
	activerecord_connect # See chapter introduction

	class BlogPost < ActiveRecord::Base
	 has_many :comments

	class Comment < ActiveRecord::Base
	 belongs_to :blog_post

Now you can create entries in the tables without writing any SQL:

	post = BlogPost.create(:title => First post,
	 :content => "Here are some pictures of our iguana.")

	comment = Comment.create(:blog_post => post, :author => Alice,
	 :content => "Thats one cute iguana!")

	post.comments.create(:author => Bob, :content => Thank you, Alice!)

You can also query the tables, relate blog posts to their comments, and relate comments back to their blog posts:

	blog_post = BlogPost.find(:first)

	puts %{#{blog_post.comments.size} comments for "#{blog_post.title}"}
	# 2 comments for "First post"

	blog_post.comments.each do |comment|
	 puts "Comment author: #{}"
	 puts "Comment: #{comment.content}"
	# Comment author: Alice
	# Comment: Thats one cute iguana!
	# Comment author: Bob
	# Comment: Thank you, Alice!

	first_comment = Comment.find(:first)
	puts %{The first comment was made on "#{first_comment.blog_post.title}"}
	# The first comment was made on "First post"


ActiveRecord uses naming conventions, database introspection, and metaprogramming to hide much of the work involved in defining a Ruby class that corresponds to a database table. All you have to do is define the classes (BlogPost and Comment, in our example) and the relationships between them (BlogPost has_many :comments, Comment belongs_to :blog_post).

Our tables are designed to fit ActiveRecords conventions about table and field names. The table names are lowercase, pluralized noun phrases, with underscores separating the words. The table names blog_posts and comments correspond to the Ruby classes BlogPost and Comment.

Also notice that each table has an autoincremented id field named id. This is a convention defined by ActiveRecord. Foreign key references are also named by convention: blog_post_id refers to the id field of the blog_posts table. Its possible to change ActiveRecords assumptions about naming, but its simpler to just design your tables to fit the default assumptions.

For "normal" columns, the ones that don participate in relationships with other tables, you don need to do anything special. ActiveRecord examines the database tables themselves to find out which columns are available. This is how we were able to use accessor methods for blog_posts.title without explicitly defining them: we defined them in the database, and ActiveRecord picked them up.

Relationships between tables are defined within Ruby code, using decorator methods. Again, naming conventions simplify the work. The call to the has_many decorator in the BlogPost definition creates a one-to-many relationship between blog posts and comments. You can then call BlogPost#comments to get an array full of comments for a particular post. The call to belongs_to in the Comment definition creates the same relationship in reverse.

There are two more decorator methods that describe relationships between tables. One of them is the has_one association, which is rarely used: if theres a one-to-one relationship between the rows in two tables, then you should probably just merge the tables.

The other decorator is has_and_belongs_to_many, which lets you join two different tables with an intermediate join table. This lets you create many-to-many relationships, common in (to take one example) permissioning systems.

For an example of has_and_belongs_to_many, lets make our blog a collaborative effort. Well add an users table to contain the posts authors names, and fix it so that each blog post can have multiple authors. Of course, each author can also contribute to multiple posts, so weve got a many-to-many relationship between users and blog posts.

	use cookbook;

	 name VARCHAR(200),

Because a blog post can have multiple authors, we can just add an author_id field to the blog_posts table. That would only give us space for a single author per blog post. Instead, we create a join table that maps authors to blog posts.

	use cookbook;

	DROP TABLE IF EXISTS blog_posts_users;
	CREATE TABLE blog_posts_users (
	 blog_post_id INT(11),
	 user_id INT(11)

Heres another naming convention. ActiveRecord expects you to name a join table with the names of the tables that it joins, concatenated together with underscores. It expects the table names to be in alphabetical order (in this case, the blog_posts table comes before the users table).

Now we can create a User class that mirrors the users table, and modify the BlogPost class to reflect its new relationship with users:

	class User < ActiveRecord::Base
	 has_and_belongs_to_many :blog_posts

	class BlogPost < ActiveRecord::Base
	 has_and_belongs_to_many :authors, :class_name => User
	 has_many :comments, :dependent => true

The has_and_belongs_to_many decorator method defines methods that navigate the join table. We specify the :class_name argument because otherwise ActiveRecord has no idea which ActiveRecord class corresponds to an "authors" relationship. Without :class_name, it would look for a nonexistent Author class.

With the relationships in place, its easy to find blog posts for an author, and authors for a blog post:

	# Retroactively make Bob and Carol the collaborative authors of our
	# first blog post.
	User.create(:name => Bob, :blog_posts => [post])
	User.create(:name => Carol, :blog_posts => [post])

	author = User.find(:first)
	puts "#{} has made #{author.blog_posts.size} blog post(s)."
	# Bob has made 1 blog post(s).

	puts %{The blog post "#{post.title}" has #{post.authors.size} author(s).}
	# The blog post "First post" has 2 author(s).

As with the has_many or belongs_to relationships, the has_and_belongs_to_many relationship gives you a create method that lets you create new items and their relationships to other items:

	author.blog_posts.create(:title => Second post,
	 :content => We have some cats as well.)

And since the blog_posts method returns an array-like object, you can iterate over it to find all the blog posts to which a given user contributed:

	author.blog_posts.each do |post|
	 puts %{#{}s blog post "#{post.title}" } +
	 "has #{post.comments.size} comments."
	# Bobs blog post "First post" has 2 comments.
	# Bobs blog post "Second post" has 0 comments.

If you want to delete an item from the database, you can use the destroy method available to all ActiveRecord objects:


However, deleting a blog post does not automatically remove all the comments associated with that blog post. You must tell ActiveRecord that comments cannot exist independently of a blog post, like so:

	class BlogPost < ActiveRecord::Base
	 has_many :comments, :dependent => destroy

Why doesn ActiveRecord do this automatically? Because its not always a good idea. Think about authors: unlike comments, authors can exist independently of a blog post. Deleting a blog post shouldn automatically delete all of its authors. ActiveRecord depends on you to make this kind of judgment, using your knowledge about your application.

See Also

  • Recipe 15.7, "Understanding Pluralization Rules," for more on the connection between the table name and the ActiveRecord class name



Date and Time



Files and Directories

Code Blocks and Iteration

Objects and Classes8

Modules and Namespaces

Reflection and Metaprogramming


Graphics and Other File Formats

Databases and Persistence

Internet Services

Web Development Ruby on Rails

Web Services and Distributed Programming

Testing, Debugging, Optimizing, and Documenting

Packaging and Distributing Software

Automating Tasks with Rake

Multitasking and Multithreading

User Interface

Extending Ruby with Other Languages

System Administration

Ruby Cookbook
Ruby Cookbook (Cookbooks (OReilly))
ISBN: 0596523696
EAN: 2147483647
Year: N/A
Pages: 399 © 2008-2020.
If you may any questions please contact us: